Links

Ben Laurie blathering


Bitcoin 2

Well, that got a flood of comments.

Suppose I take 20 £5 notes, burn them and offer you a certificate for the smoke for £101. Would you buy the certificate?

This is the value proposition of Bitcoin. I don’t get it. How does that make sense? Why would you burn £100 worth of non-renewable resources and then use it to represent £100 of buying power. Really? That’s just nuts, isn’t it?

I mean, it’s nice for the early adopters, so long as new suckers keep coming along. But in the long run it’s just a pointless waste of stuff we can never get back.

Secondly, the point of referencing “Proof-of-work Proves Not to Work” was just to highlight that cycles are much cheaper for some people than others (particularly botnet operators), which makes them a poor fit for defence.

Finally, consensus is easy if the majority are honest. And then coins become cheap to make. Just saying.

17 Comments »

  1. Following their analogy to mining, it seems like your argument would also apply to burning fossil fuel to extract gold from the ground and transport it to vaults. In both cases, a non-renewable resource is used up to produce receipts for work done. Not saying gold-based money is an ideal, but it did work for quite some time and in some ways still does.

    Your second point also seems to fit the analogy, since gold is more abundant and cheaper to extract in different parts of the world and at different points in history. The owners of this cheap gold seem to be lucky in the same way that early adopters are lucky.

    Comment by Tyler Close — 20 May 2011 @ 17:12

  2. Dear Ben:

    I’m rather surprised at you posting this and your previous entry. You’ve either failed to understand BitCoin before posting, or else you can envision a radical improvement on BitCoin which I can’t. I sure hope it is the latter! ;-)

    So: how do we allocate money in our new successor-to-BitCoin decentralized currency without wasting CPU cycles? By Coase’s theorem, it doesn’t actually matter *that* much how we do it, but I’d like to choose a way that entices a wide group of people to become stakeholders. (My first idea was simply that I would assign all of the money to myself and that I would distribute it philanthropically to whomever I wanted. However, I suspect this wouldn’t yield acceptance as widespread as BitCoin’s…)

    Keep in mind that this question is really only about the *initial allocation* of the money. Once this phase has passed there will be no further creation of money.

    Okay, once we’ve solved the initial-allocation problem without wasting energy, then we can turn our attention to the consensus-building problem without wasting energy. BitCoin effectively uses hashcash as a defense against Sybil attack–to control the consensus you have to not only have a majority of the “identities” or IP addresses or whatever, but you also have the have a majority of the CPU power. Maybe we could think of a better way to defend against that attack. Let’s do it! :-)

    I do hope you have carefully read http://www.vsewiki.cz/images/archive/8/89/20110124151146!Bitcoin.pdf already and if you haven’t that you do so before you post again.

    Regards,

    Zooko

    Comment by Zooko — 20 May 2011 @ 17:20

  3. Better link to the paper: http://www.bitcoin.org/bitcoin.pdf

    Comment by Zooko — 20 May 2011 @ 17:21

  4. Ben – your instinctive analysis of Bitcoin seems spot on, unsurprisingly. On the other hand, little has changed technologically on the internet since 1999, and it hasn’t prevented unjustified waves of enthusiasm for all sorts of enterprises, from Twitter onward.

    People are going to like what they like, and there’s no accounting for taste. Bitcoin, like the recent commercial phenomenon Groupon, tends to turn people into marketers because they feel they have something to gain, however small it might be in the end; I think that partly accounts for its temporary success. I suspect, and have even come to hope, that the enthusiasm will be fleeting as people recognize it’s functionally a scam (even if it has the potential for more in concept, reworked a bit).

    I’ve written up more detailed analysis of Bitcoin’s functional problems (and notes about what it could technically offer in theory in another form) here, in case they’re useful to you in any way:

    http://www.fatwallet.com/forums/finance/1090435/m15945892/#m15945892
    http://www.bogleheads.org/forum/viewtopic.php?t=74834

    I’ve done it pseudonymously for now because the rage of the nutters on the official Bitcoin forum is a bit frightening. “I reject all law and morality” is a common sentiment there, and the most prominent of the early adopters are rather bizarre. One lives in the US midwest and, as a hedge against the apocalypse, hoards guns, ammo, and silver — the latter to disinfect his raw milk when society collapses. They generally follow Mises and Ayn Rand uncritically, prize selfishness, live by a maxim of caveat emptor, and so on. Or at least the loudest do. It’s created a sort of groupthink in which people are promoting the idea that a single Bitcoin will be worth hundreds or thousands of US dollars, and in this in turn functionally perpetrates a fraud on the public.

    Comment by ComputerScientist — 20 May 2011 @ 17:32

  5. What if I chopped down a tree, mulched it into pulp, dried it into paper and stamped some ink on it, handed it to you and told you it was worth £100. Really? That’s just nuts, isn’t it? That’s current exactly what all the central banks are doing.

    Comment by John Xovox — 20 May 2011 @ 18:21

  6. @John Xovox, the paper note isn’t worth anything. It is an agreement, saying that the US government or the bank of england owes you some actual money. Nobody has ever seen this actual money. :-)

    Comment by Luke Palmer — 20 May 2011 @ 19:51

  7. “Suppose I take 20 £5 notes, burn them and offer you a certificate for the smoke for £101. Would you buy the certificate?”

    Yes, if I have confidence that others will be happy to accept the certificate in exchange for goods or services of equivalent value. That’s what cash is.

    You seem to be focusing on the cost of running the protocol. The interesting bit about bitcoin is that the protocol is just a means to an end — a fiat currency guaranteed by the self interests of a large number of distributed participants, rather than a centralized bank. The costs of running a Bitcoin network seem to be of the same magnitude as, and possibly a few multiples lower than, the costs of running a centralized bank.

    I’m very skeptical about bitcoin, but your criticisms seem wide of the mark so far.

    Comment by Andy — 20 May 2011 @ 21:08

  8. I agree w/ ComputerScientist on basically everything except the argument about hiding
    data. That was dumb for reasons mentioned (can encode same stuff in blogger, twitter, or
    imageshack in a much denser manner). He’s also absolutely on-point about the forum being
    a huge turnoff to people. I’m a grad student in CS (crypto’s not my area of expertise,
    but I try to read crypto conf proceedings to keep current), and I would be embarassed to
    associate myself with everyone on that forum except maybe 4 people. I’ve talked about it
    with a couple friends/coworkers in the dept here who I know well, but I would be
    embarassed to point “polite company” anywhere near the bitcoin forums.

    I think bitcoin is a good idea modulo some scalability issues that seem fixable. But I
    think the main thing to take away from all this is to realize the following: even with
    all the flaws you can think of, people are so DESPERATE for the ability to conduct
    LOW-COST, UNDENIABLE/UNREFUTABLE transactions over the internet that they are willing to
    PAY REAL USD for these weird “bitcoins” on the HOPE it might do what they need. Stop and
    read that again. Clearly, something is very very wrong with the world when the obstacles
    or regulations towards someone creating a company to do this are so great that noone has
    done it (it seems there’s relatively new company, Dwolla, charging 0.25 per transaction.
    I’d personally like to see < 0.01, which should still remain profitable).

    I too think such a system is crucially needed. My main worry with bitcoin is the
    anonymity of the creator and the fact that he owns a large % of the economy. The cynic
    deep inside of me tells me that bitcoin was started by a brilliant computer scientist who
    happened to have control of a massive botnet as a way to monetize it. Even if that's not
    the case, the "one vote per CPU" model makes botnet herders king, and I don't want an
    economic system like that.

    The repeated-hashing is a very clever way of doing byzantine agreement without massive
    amounts of communication (e.g. n^2 communication for standard byzantine agreement
    protocol). I had not seen this done before bitcoin, nor in any of the references of
    "Satoshi"'s paper, so the idea of trading communication for CPU time in a byzantine
    agreement protocol seems to be a novel, clever, idea of his. Can we do byzantine
    agreement with even less?

    Also, the idea of having bitcoins be assigned to autonomous programs/scripts is quite
    clever too, and allows for some amazing things automatically like coins that can be spent
    by k-out-of-n people. With a system like this, one could create
    all kinds of information markets without ANY escrow service needed (two parties can
    exchange the "key" needed to spend a coin for some other information, like an invert to a
    hash, by doing the exchange bit-by-bit and after each bit swap, proving in zero-knowledge
    that there is a secret agreeing with the bits sent so far).

    Comment by AnotherComputerScientist — 21 May 2011 @ 3:39

  9. PseudoScientist said: “They generally follow Mises and Ayn Rand uncritically…”

    BS alert. The two are mutually exclusive. Followers of one really dislike the followers of the other.

    You just proved, as in proof, that you talk out your posterior.

    Comment by JD — 21 May 2011 @ 8:12

  10. To commenter #7: Thanks for the general agreement and thoughts.

    I probably should have made my point about steganography in the block chain clearer. I only meant that it invited regulatory attention and perhaps imposed a public barrier to adoption (if the block chain comes to be used to store arbitrary data). To put the point more narrowly, a technology that requires a consensus-based development of a fully replicated, potentially permanent database (because Satoshi’s “pruning” can’t apply to unspent coins) invites difficulty when an attacker can include whatever he’d or she’d like in it, drawing whatever attention or opposition the attacker wants to draw. Twitter doesn’t face this problem as an entity because it can remove whatever content an attacker adds; that’s not an option for Bitcoin without some out-of-band protocol for achieving meta-consensus as to what counts as illegitimate data for the block chain. Imagine the (legal and social) problem for Twitter if it had to accept what everyone posted with no possibility of removing it, and that’s the problem I was saying Bitcoin potentially faces.

    I agree with you that technology ought to be able to improve financial services. I’m just afraid that the empirics of the matter are complicated, and they don’t seem to favor Bitcoin. For example, I’m not confident in the steady state that Bitcoin could beat Dwolla’s transaction fee. Already Bitcoin essentially costs 6 or 7 cents per transaction, and there are many circumstances in which that number will rise — particularly as the block-creation subsidy begins to decrease. And that’s just within the Bitcoin system itself, not counting conversion fees between other currencies, the downsides of volatility, the likelihood of fraud (which has already manifested itself among currency converters), and so on. Of the three or four sensitive commentators in the Bitcoin forums that I’ve encountered, I think more than half question Bitcoin’s ability to be competitive on transaction fees in the long run. Someone has to pay for all the energy, oversight, etc., even after the block subsidies vanish. Absolute decentralization has significant costs.

    Here’s another way to look at it: Bitcoin tries to optimize for anonymity, Austrian-school monetary policy, and decentralization of the money-issuing function. That all has a significant cost, and it seems likely that a system that doesn’t optimize for those features simply wouldn’t cost as much. (And that’s even putting aside the very high seigniorage to Satoshi and the early adopters that you mention.) I very strongly doubt that most people would want to pay extra, in every transaction, for those features — even if slightly more financial privacy would be a good thing on balance. Bitcoin thus seems, on a rational rather than a speculative analysis, to be designed functionally as a niche product for those who want, legitimately or otherwise, to engage in very private financial transactions at very significant cost. That seems to boil it down nicely; Bitcoin is that plus a get-rich-quick scheme for the early adopters.

    Comment by ComputerScientist — 21 May 2011 @ 8:20

  11. (Sorry! I meant “To commenter #8.”)

    Comment by ComputerScientist — 21 May 2011 @ 8:21

  12. The thing you need to understand about bitcoin is that it’s explicitly designed with the Austrian understanding of the value of money in mind (Satoshi was clearly a libertarian, and bitcoin is designed in a way to mimic precious metals as a medium of exchange). With this understanding, money SHOULD be difficult to produce. For instance, if alchemy were cheap and possible it would still be decentralised production but gold would be useless as a medium of exchange. The hard limit of 21 million BTC is like saying that it is the total amount of gold in the ground, and eventually it will all be found. And yes there will be a gold rush first, and it will be easy to come across gold with a simple pan, but eventually you’ll have to save and invest into mining equipment to dig deep in order to find gold and eventually even that won’t matter anymore. In a similar way, yes the gold panners were lucky early adopters it’s not a problem. The wealth gap isn’t actually an issue because wealth isn’t a fixed pie anyway.

    The supply of gold is limited; you don’t want heavy inflation. That is why paper fiat money is dangerous, because it’s cheap for the government to produce and the demand is created because of legal tender laws (i.e. the initiation of force). If you don’t think this is the case, advocate for the repeal of legal tender laws and let’s see happens. Fiat paper money is an easy way for the government to tax people in a hidden way; inflation transfers purchasing power to the person who produces the money. In the case of gold, it’s difficult and decentralised to produce, but in the case of fiat paper money, it’s centralised and cheap to produce with demand kept high through the means of their so-called right to initiate force.

    In a sense you are not attacking bitcoin per se but the Austrian school. If you want to attack the economics behind this understanding of the value of money, read some of the vast amount of free literature at mises.org on this topic (it’s a fantastic resource) and tell us all why it’s nonsense. But this knee jerk reaction of yours is not a compelling one.

    Comment by Jack — 21 May 2011 @ 9:38

  13. “The cynic deep inside of me tells me that bitcoin was started by a brilliant computer scientist who happened to have control of a massive botnet as a way to monetize it. Even if that’s not the case, the “one vote per CPU” model makes botnet herders king, and I don’t want an economic system like that.”

    you can clearly see that was not the case by simply checking how network computation power risen – very slowly

    cpu is basically useless for mining as it is hundreds times slower than single ati 5870 card that is most widely used by miners, even if botnets start mining it changes nothing – someone just would have some bitcoins

    “Imagine the (legal and social) problem for Twitter if it had to accept what everyone posted with no possibility of removing it, and that’s the problem I was saying Bitcoin potentially faces”

    even if there is some ‘bad’ data in blockchain it is not that different than sending bank transfer with ‘Joe is dick’ as description… or drawing some dicks on dollars notes… hardly a problem for me anyway as i don’t feel the urge of staring into that particular file in search for something that may offend me

    “I’ve done it pseudonymously for now because the rage of the nutters on the official Bitcoin forum is a bit frightening. “I reject all law and morality” is a common sentiment there”

    that kind of statement is just pure nonsense, you fail to understand whole concept of btc and simply going ad personam… are you politician or something? ;)

    Comment by bye — 21 May 2011 @ 15:14

  14. “Already Bitcoin essentially costs 6 or 7 cents per transaction”

    Not true.

    “official Bitcoin forum is a bit frightening. “I reject all law and morality” is a common sentiment there””

    Well, there is bunch of hippies there, but how that changes nature of bitcoin itself?

    “My main worry with bitcoin is the anonymity of the creator and the fact that he owns a large % of the economy.”

    It is always the case. If there is any project going, there is always someone behind it. Period.
    There is nothing that anyone can change in bitcoins except exchange rate – THIS is it’s strength.

    Comment by sir Edward — 21 May 2011 @ 15:29

  15. If you gave me 5 $20 bills and I spent them traveling to California in 1840, would you accept $101 worth of a somewhat rare metal that I dug out of the ground as repayment? Why? Did I burn those $20 bills or did I spend them for goods and services? Does $20 of computing time get burned or does it pay engineers, coal miners, etc.?

    Of course gold has utility not just as a monetary standard. Bitcoins only have the utility of proving I “went to the gold rush”, but the monetary use of gold is quite similar to bitcoins. “Real” money is spent chasing after metal in the ground which then becomes “real” money mostly because of the “real” cost of mining gold.

    Comment by Ben — 21 May 2011 @ 18:14

  16. I like Bitcoin and try to use it. I am reasonably intelligent and as well versed in crypto as any computer geek. Could you please try to phrase your critique of Bitcoin in a way that I can understand?

    Yes, I did read the paper you likned. Yes, I’m casually aware of Digicash and other early crypto cash. They did not interest me as they either used credit card transactions (and the economic transaction monopoly is what I would like to see changed to a distributed model instead), or their inventors had the brilliant idea that they owned all the money from the start (which is a system that nobody else should take interest in).

    Your critique so far is some hand waving about smoking money? What? Please. Bitcoins are already in small use. Please do us a favour and treat us as intelligent and describe what properties could be improved.

    Comment by Jonas — 21 May 2011 @ 22:24

  17. If there’s an exchange up and running where those smoke certificates are trading for, say 120 lb, then sure, I’ll buy yours for 101 lb.

    Comment by Michael Kirkland — 25 Jul 2011 @ 17:10

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress

Close
E-mail It